In this study, a deep learning model is developed to classify brain tumor types using MRI images. Our model uses Convolutional Neural Networks (CNNs) combined with recurrent layers to enhance classification performance. The architecture includes several convolutional layers for feature extraction, followed by LSTM and BiLSTM layers to capture sequential patterns in the data. This hybrid strategy is supplemented with data augmentation strategies that boost model generalization and reduce overfitting. The dataset, which consists of greyscale MRI scans classified into four categories (Glioma, Meningioma, Notumor, and Pituitary), is preprocessed and divided into training and test sets. The model’s performance is measured using accuracy and a confusion matrix, demonstrating its ability to correctly identify tumors. Misclassified samples and predictions on new images are also analyzed, demonstrating the model’s practical applicability in medical imaging.